A number of confluent lines of evidence support a role for REM sleep in learning and memory consolidation. In the first grant period we showed changes in hippocampal neural activity during REM that correlate with experience and follow the course of memory consolidation (SA1), namely a reversal in phase of firing from theta rhythm peaks to troughs. We altered learning rates and performance strategies with REM deprivation (SA2) and enhancement (SA3) on a new behavioral task and measured the changes in the theta peak-to-trough reversal with these manipulations. Instead of manipulating REM sleep, this proposal seeks to determine the physiological events of REM sleep that are important for memory consolidation in a hippocampus-dependent task. We hypothesize that the absence of norepinephrine and serotonin during hippocampal sleep reactivation provides an environment wherein synapses in the direct entorhinal-to-CA1 path may be strengthened and CA3-to-CA1 synapses may be weakened (depotentiated). Depotentiation is theoretically necessary for clearing the temporary, modifiable associative memory network of the hippocampus to allow changes in the memory code with the arrival of new information. In Specific Aim 1 we will test the necessity of hippocampal reactivation during sleep for depotentiation and test the necessity of depotentiation for successful modification of memory in a switching task. By allowing sleep to occur normally while temporarily inactivating the hippocampus, we will isolate the contribution of hippocampal reactivation at fixed time points in the memory consolidation process. We expect that the mnemonic benefits of sleep will not be realized in hippocampal-dependent tasks when the hippocampus is inactivated before consolidation is complete. When we release sleep reactivation, we expect to see neural activity patterns in waking and in REM sleep that indicate the consolidation process had been arrested and the peak-to-trough theta phase reversal delayed. We also expect consolidation to continue normally from the point at which it had been stopped as shown by resumed improvement in behavioral performance. In Specific Aims 2 and 3 we will test the mnemonic necessity of the absence of the two neurotransmitter systems normally silent during REM sleep. We will allow REM sleep to occur, but introduce noradrenergic and serotoniergic agonists in the hippocampus during sleep, measuring the changes in learning and in the sleep/waking profile of cellular activity as in the first specific aim. These results should suggest mnemonic influences of different antidepressants which act with varying efficacy on the monoaminergic axis. In addition, this effort will contribute to the knowledge base of memory consolidation mechanisms to ultimately indicate effective ways to improve the memory and learning capabilities of impaired individuals.